A Hacker News for grad students?

[dcy]

I love hacker news. The community . The unsaid agreement to uphold all that is clever and awesome. The shared intelligence of a stimulating community. But I often find myself wishing for a little more (Is there no limit to human desire). Don't get me wrong I've learned more here than anywhere else. But the relentless noise of software often drowns out interesting posts in math, physics and applied sciences . Maybe someplace where grad students and their ilk may share their experiments , thoughts , thought-experiments and engage in futile discussions on the definition of 'insight' in mathematics . Comment below with your thoughts.

[j2kun]

I am the author of a blog called Math Intersect Programming (and a graduate student in CS theory), and I also would like such a community.

I'm not happy with datatau because I feel it's too focused on basic software tutorials. I'm not happy with any of the subreddits I've looked at; I think a large part of it is just that by being part of reddit the quality of discussion is pulled down.

In more generality, I feel that people who want to engage in scientific discourse at the graduate student level (I'm thinking of mathematics, physics, and computer science) don't have enough tools to do this. There are conferences, which are typically stuck in one field; and blogs by researchers, which are great but often do too much conference/workshop/program advertising for my taste. Most web spaces like reddit are too elementary (how many posts do I read there about picture proofs of sums of squares!). I think the subcommunity of HN users interested in mathematics, and the kinds of posts that end up on HN, shows that there is real interest. I even know of mathematics/CS researchers who keep an eye on HN because of the promptness of dissemination and quality of discussion.

All this being said, I would be very willing to help run such a community, getting it off the ground with interesting submissions and such. Right now the closest thing I have is a Google Plus community I've formed around my blog [1]. But again, very few people beside me post things that I consider up to my standards.

[1]: plus.google.com/communities/101551468332631556735

[chaoxu → j2kun]

Also a graduate student in cs theory and was a BSMer(huh, I realized I attended your midwest theory day talk)

There is a facebook group, computer science real talk. https://www.facebook.com/groups/500102603385910/ Things in there are interesting to grad students. It seems to start as a place for cornell cs graduate students. There are people talking about STOC papers in there, neat.

[hoonose → chaoxu]

Thanks for sharing! I founded the group since I was fed up with the Computer Science Facebook group for Cornell -- while there was occasionally interesting CS content, it was flooded by a lot of stuff that's either no longer interesting to me (a lot of it is Cornell undergraduate centric, while I'm now a graduate student at MIT), or just meme/jokish in nature. My goal was to make a group with interesting, thought-provoking content and better quality discussion. If this appeals to you, feel free to join and contribute.

[j2kun → chaoxu]

Sounds like an awesome resource! I was also a BSMer :)

[JacobEdelman → j2kun]

I'd also like to thank you for your blog. It is a resource that I have repeatedly found myself going back to as I explore CS and math.

[jcurbo → j2kun]

I'd just like to say thanks for your blog, I find the approach you take very appealing and I especially like the primers.

[EvanMiller]

I would also like to see such a community.

I've found it somewhat in the Julia community, which tends to lean heavily towards scientific computing and has a lot of Ph.D.s, grad students, and Ph.D. dropouts. We just had JuliaCon last week, and the talks tended to be first about a problem domain (statistical modeling, optimization, natural language processing, finite element methods...) and secondarily about Julia. I loved it.

If you're in the Chicago area, I'd encourage you to drop by our Julia Meetup group, which has a similar format. Previous talks have been about solving problems in climate modeling, machine learning, and molecular dynamics, and I'm trying to line up talks later this summer about machine learning in psychology and MLE modeling of longitudinal econometric data. (All using Julia, of course.) The group is small, but the quality of discussion is very high. Join the group and you'll get an email when the next meeting is added to the calendar:

http://www.meetup.com/JuliaChicago/

[sktrdie]

I find Academia.StackExchanage [1] and the entire StackExchange ecosystem [2] quite awesome for learning, teaching and simply discussing about anything related to specific communities.

For anything else there's always Mailing Lists (ML). What I've learned over the years is that it doesn't really matter which technology or which format is used to have a discussion online (IRC, Mailing Lists, Reddit/HN sites). What matters is the community and the people behind it.

For example on IRC, specifically freenode, there's tons of channels with tremendous quality of discourse because of the moderators being able to keep the quality that way.

If you're looking for discourse about Math, try #math on freenode and signup to math.stackexchange and see which mailing lists exist out there. Perhaps even look into Math academic journals out there and see whether there are mailing lists to start discussing things.

1. http://academia.stackexchange.com/

2. http://stackexchange.com/sites

[j2kun → sktrdie]

This is a good point. I have found cstheory.stackexchange and math.stackexchange a good place to field questions of mine (with the latter mostly being for stupid questions that I should already know the answer to).

That being said, the stackexchange community was designed for the purpose of quashing discussion, especially that which doesn't directly address the question at hand, even if those discussions are productive and insightful.

[DanAndersen]

I don't really know of any spaces like that (and as a CS PhD student I would enjoy something like HN for grad students), but just to latch onto this general post -- does anyone have links to good Hacker-News-like sites for various topics? There's a lot I like about HN but the whole startup scene and "yet-another-web-framework.js" is less relevant to me.

[Curmudgel → DanAndersen]

You'll probably like Lambda The Ultimate. It's a Programming Languages weblog with discussions and forums, but their traffic and updates are miniscule in comparison to news.ycombinator. If you're a busy PhD student, that's probably not a bad thing.

http://lambda-the-ultimate.org/

[izyda → DanAndersen]

This site has a lot of promise: http://www.datatau.com/ Unfortunately, the community is not yet large enough or active enough to have a ton of new posts everyday. That said, there are many quality links

[Sambdala → DanAndersen]

http://hn4hn4x.herokuapp.com

[compedit → DanAndersen]

Software wise, there is https://lobste.rs/.

It's updated fairly often, (the oldest link on the front page is only 3 days old), and the comments are typically high quality.

It is invite only though atm.

[natural219 → DanAndersen]

We're all searching, brother.

[seltzered_]

I'm trying to start such a thing with a post-grad friend with http://www.journaltalk.org. It's currently a clone of lobste.rs (an hn clone with web-of-trust invite), but we want to start a community where people talk about articles and they'll be tagged by journal, similar to 'journal clubs' grad students and phd's have. If you email me@vivekgani.com I'll give you an invite.

EDIT: just noticed the servers down right now, eek! I'll look into it later tonight. we also keep source (again, essentially just a lobste.rs fork for now) at https://github.com/seltzered/journaltalk

[jcurbo → seltzered_]

Sounds interesting! I like the idea of organizing it by paper/journal/article. It could be an online version of a Papers We Love meetup.

[abdullahkhalids]

Of all the possible HN alternates that have been proposed, this one is probably the one that makes sense. I don't think there is any alternate at the moment. I have a couple of thoughts and queries.

1. You put grad students in the title. But you only mention natural sciences and mathematics in the text. What about social science? What about the humanities? I am a physicist but I would personally like to know about progress on major problems in all academic fields. If you limit yourself to only some disciplines how will draw the line and why? HN has the line "interesting to hackers", which makes it very very broad. It relies on undemocratic voting, and occasionally moderation, to separate the good from the bad.

2. The reason the quality of arguments at HN is relatively high, is primarily the software behind the HN scores system. It very effectively kills of many techniques that are used by people to raise their scores. Arguably, getting on HN front page is directly correlated to the visits/money you get, while on an academic forum people will only care for egoistical reasons. The problem will be less, but it will still exist. If the community takes on and starts to grow, it will need access to the HN points software which is not public at the moment as far as I know.

3. Discussion surrounding links to papers, news articles or essays are generally serious and 'valuable'. This is what HN does primarily. You are proposing a forum where people can talk about their own thoughts and ideas. There are many such internet forums for this purpose. Most of them degenerate into homework problem threads and how to get into grad school questions. Why? Because what you thought over breakfast this morning is usually not brilliant enough, while a paper/essay/news has had many hours of professional thought behind it.

4. It might be possible to piggyback over HN. Any link on HN relating to grad students can be manually resubmitted on the community website. The discussion happens on HN. The community website only serves as a curated list. This will solve the problem of drowning in the noise of software. Of course, the limitation will be that all posts submitted will have to conform to the "interest to hackers" criteria.

[Mz → abdullahkhalids]

The reason the quality of arguments at HN is relatively high, is primarily the software behind the HN scores system.

This observation drastically undervalues the fact that this is a business space and part of the business model for Y Combinator, so for some people, participation here is significant to their career and potentially worth millions of dollars. It cuts way back on the assholery. Plus, since it is part of a business model, the forum owners work pretty hard on the social piece. There is, no doubt, still room for improvement but my experience has been that forums that are part of a business model are generally dramatically better than those run for free by idealists who frequently get really crabby about being underappreciated, the forum turning into too much work, etc.

[001sky → Mz]

It'm not sure its the business model, but that it is attached to an enourmous pile of assets/cash. That might be parsing your comment a bit, but the distinction is both relevant and important if you are trying to replicate it.

[Mz → 001sky]

There are other business models that work (like MetaFilter -- very different model but still a business). HN is part of the Y Combinator business model in that you must have (last I checked) a Hacker News handle to fill out the application for Y Combinator. Every founder has to have their own handle. So it is a funnel and things you say here are kind of de facto part of the application. Thus, being a jackass here on a routine basis can hurt your chances of getting in. For other people, who have no plans to apply to Y Combinator, there is still, yes, a lot of money of the table for many people. Also, because it is part of their business model, they now have a full time paid moderator, something boards done out of the goodness' of someone's heart cannot afford. The difference in quality shows.

[j2kun → Mz]

I think a good thing to hope for in starting the proposed community would be that enough researchers and graduate students pay attention that being a jackass could affect your academic career in similar ways (less likely to get a postdoc or grant, etc., if someone on the committee knows your reputation).

[dcy → abdullahkhalids]

1.'Grad students' is a misnomer. I should've known better.

2.Ideally HN would've been the internet utopia that the intelligentsia turned to for respite and nourishment. But if HN(algorithm+users) performs a filtering function on all the posts submitted, then it has evolved to consistently rank posts relating to web-based software development over anything else. Now this previous sentence is rather harsh. There are many wonderful articles on physics , social sciences , philosophy, history etc that are featured on a daily basis. But for each such article upvoted there are a multitude of github.io's .(Which,again is great. HN is after all for hackers and entrepreneurs). Every now and then though I come across a post linking to a blog that provides such delightful insight , a visualization of vectors , a new perspective on chemical analysis ,a novel line of questioning on the usefulness of reason and logic that I wonder if there is a multitude of such material 'out there' that I may be missing out on.

3.By grad student I meant anyone pursuing a rigorous(again dodgy semantics) scientific education(whether enrolled or on their own) while trying to fit the pieces(the various dependencies,if you will) together.

4.As for the whole possibility of creating such a site , I was merely checking(at least at the time of posting) if I was the only one that felt the necessity .

[ivan_ah → dcy]

> fit the pieces(the various dependencies) together

Something that would work is a platform for writing review papers collaboratively. I'm imagining a review would be a mix of 30% original content (written in blog-post language) and 70% links to research articles. In the end, the results could be publishable as a review paper in a journal, or exported as an annotated bibliography (a .bib file, e.g. [1]).

[1] I started a review paper on github last year, but I haven't touched it since, will do this summer https://github.com/ivanistheone/LDAreview/

[X4 → abdullahkhalids]

In the beginning, before there were non-students, Facebook was about solving that issue, wasn't it? I'd also heartly appreciate such a platform. Most stuff I find really interesting won't be interesting at all for HN, thus I end up sharing less.

It's not a social network scientists need, but a kind of organized knowedge-rating, sharing and curation site.

[j2kun → abdullahkhalids]

I agree on (3) that such a community should mostly be restricted to things not thought of over breakfast. If it comes in the form of a submitted link to a blog post, on the other hand, that should be fair game even if the blog post author just "thought of it over breakfast."

[imurray]

Someone did briefly run an "Academic Hacker News" about 5 years ago (on HN's arc software).

https://news.ycombinator.com/item?id=506064

http://web.archive.org/web/20100107015442/http://hnacademic....

http://web.archive.org/web/20100929184746/http://www.cs.toro...

[j2kun → imurray]

It looks like this page focused mostly on posting introductory textbooks, and not so much on discourse. If you used this site when it was alive, did you feel this was characteristic of the site? (and not just a particular snapshot that happened to have the front page full of books)

[imurray → j2kun]

The intention (3rd link) was for papers and discussion. There were probably more comments in the first few days before interest fizzled. Gaining critical mass is hard.

[j2kun → imurray]

I think a site organized around "links to papers" would have an even harder time gaining critical mass because people find out about papers in so many ways (word of mouth, conferences, blogs, arXiv), and the "best" ways to find out about papers that you'd then devote a long time to reading usually need serious motivation (great talk, met an author, know you need to learn a technique) rather than a random link.

[ihnorton]

ResearchGate's Open Review [1] and PubPeer [2] have some of what you might want. Both had excellent discussion/criticism about the acid-bath stem cell paper a few months back, for example. However, at least for PubPeer, the active papers seem to be mostly biomedical.

[1] http://www.researchgate.net/publication/259984904_Stimulus-t... [2] https://pubpeer.com/publications/F0CFE0360002C25DC0BEFE28987...

[dave809]

I'd recommend LessWrong [1], It's an active community around discussing rationality, AI, self improvement, and many more intellectually stimulating topics.

[1] http://lesswrong.com/

[skadamat]

DataTau is what I recommend. Even though there's a lot of posting of how to learn basic data science and so forth, occasionally papers from ArXiv on physics and math concepts show up. End of the day, the content that shows up is a function of the community, which is small. Since it is small, it's not difficult to influence the type of content posted. I guarantee you if you and a few colleagues start posting more about research related topics, they will surface well.

[dcy → skadamat]

I agree. The site does seem a little one-dimensional but (assuming the moderators are okay with it) I can see it grow into a "hacker news for grad students" .

[meritt]

I'd try some specific subreddits such as http://www.reddit.com/r/mathematics/ or http://www.reddit.com/r/physics or http://www.reddit.com/r/AskScienceDiscussion

[sideproject]

Hi dcy.

We're currently creating a small tool that helps people create online communities. We are currently pitching it as "create your own hacker news" and we'd love for you to try it out if you have some time.

bit.ly/1r2KPVR

There are some rough edges (of course) and we're still doing our private beta at the moment, but if you sign up we would love to see how we can help!

[turnersr]

Check out Ask HN: What are some alternatives to HN? - https://news.ycombinator.com/item?id=7254884

All the subreddits listed below were aggregated into http://www.reddit.com/user/Widdershiny/m/hackernews

http://www.reddit.com/r/REMath/

http://www.reddit.com/r/ReverseEngineering/

http://www.reddit.com/r/lowlevel/

http://www.reddit.com/r/vrd/

http://www.reddit.com/r/systems/

http://www.reddit.com/r/math/

http://www.reddit.com/r/types/

http://www.reddit.com/r/dependent_types/

http://www.reddit.com/r/Coq/

http://www.reddit.com/r/haskell/

http://www.reddit.com/r/ocaml/

http://www.reddit.com/r/crypto/

http://www.reddit.com/r/fsharp/

http://www.reddit.com/r/scala/

http://www.reddit.com/r/lisp/

http://www.reddit.com/r/ProgrammingLanguages/

http://www.reddit.com/r/dataisbeautiful/

http://www.reddit.com/r/PhilosophyofMath/

http://www.reddit.com/r/AcademicPhilosophy/

http://www.reddit.com/r/PhilosophyofScience/

http://www.reddit.com/r/MachineLearning/

http://www.reddit.com/r/IPython/

http://www.reddit.com/r/statistics/

http://www.reddit.com/r/Design/

http://www.reddit.com/r/museum/

[Mz]

I know of at least one PHD student who seems fond of this site: http://linkstothedamnpaper.org/

On the about page, it describes itself as an open discussion community showcasing the best in freely-available biology research.

[bfwi]

It's possible to create your own Hacker News/Reddit clone using Telescope (http://telesc.pe/), which is built on Meteor. Of course the challenge will be to build the community that makes such a site useful.

[coulomb]

Try researchfields.com -- HN-like site dedicated to researchers and grad students for discussing that sort of thing.

[return0]

I would suggest you take a look at http://sciboards.org

[untilHellbanned]

Onarbor, https://onarbor.com, is exactly this.

[meldanpete22 → untilHellbanned]

Onarbor does have "Sites" that are representative of all the major grad schools departments, e.g., Biology, https://onarbor.com/sites/biology